Something went wrong with cudnn?

Downloaded from the official website, reinstalled cudnn, and tried using conda, but it didn’t work.

Traceback (most recent call last):
File “D:\workspace\rl\hebbrl\hebbrl3cudacopy2.py”, line 101, in
train(loader)
File “D:\workspace\rl\hebbrl\AutoEncoder.py”, line 96, in train
loss.backward()
File “D:\Users\miaoy\anaconda3\lib\site-packages\torch_tensor.py”, line 396, in backward
torch.autograd.backward(self, gradient, retain_graph, create_graph, inputs=inputs)
File “D:\Users\miaoy\anaconda3\lib\site-packages\torch\autograd_init_.py”, line 173, in backward
Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass
RuntimeError: Unable to find a valid cuDNN algorithm to run convolution

Are you running out of memory? If not, could you post a minimal executable code snippet to reproduce the issue as well as the output of python -m torch.utils.collect-env, please?

image
I try to record the memory in gpu-z.

image
my gpu is 2060 with 6g memory.